PROJECT SUMMARY Adults with developmental disabilities die at ages approximately 20 years younger than peers in the general U.S. population. This striking mortality disadvantage is larger than the mortality gap between men and women, Whites and Blacks, and adults in the top and bottom income quartiles. The reasons for the disparity are poorly understood despite decades of research on the topic. However, new research shows that a rampant practice of erroneously coding death certificates for adults with developmental disability results in widespread misclassification of their underlying causes of death. Furthermore, these erroneous codes obscure alarmingly high rates of death due to choking and Alzheimer?s disease and related dementias. While uncovering the obscuring effect of erroneous coding was a crucial advance in research, ongoing efforts to eliminate this misguided practice require empirical evidence on the multilevel factors that determine erroneous coding. The proposed project will systematically assess whether erroneous coding disproportionately occurs (1) among certain sociodemographic subgroups of decedents, (2) when certain comorbid diseases or injuries are present at the time of death, and (3) across certain U.S. state contexts. By identifying which subgroups are at greatest risk of misclassification, results will inform strategies to prevent misclassification and, thereby, determine robust estimates of cause-specific mortality among these adults. The study?s central hypothesis is that erroneous codes are more commonly found among decedents from marginalized social groups, in situations when the true cause of death was highly preventable (e.g., choking), and within U.S. states where fiscal and legal policies are less amenable to the needs of individuals with developmental disability. Using data from the National Vital Statistics System 2012?2016 U.S. Multiple Cause-of-Death Mortality files merged with state-level information on financial commitments to developmental disability services and laws governing the certification of death certificates, this project will address three Specific Aims: (1) Examine the extent to which decedents? sociodemographic characteristics predict the likelihood of having developmental disability erroneously coded as the underlying cause of death; (2) Assess whether the presence of comorbid diseases and injuries present at the time of death, with an emphasis on choking incidents, predicts the likelihood that a developmental disability is erroneously coded as underlying cause of death; and (3) Determine whether variation in funding for developmental disability services among states, and in state laws regarding who can certify a death certificate, predict whether developmental disability is erroneously coded as underlying cause of death. Results will provide a multilevel understanding of how individual-level sociodemographic characteristics, individual-level comorbid diseases and injuries, and state-level policies determine erroneous coding among adults with developmental disabilities. Based on these results, the project will present strategies for preventing misclassification. The long-term goal of this project is to better understand and reduce mortality among this historically disadvantaged population.